As the camera related technology (For example, the sensing technology of the sensing device and the image processing technology of processing device, etc.) moves forwards, the camera nowadays can, under all kinds of environments (For example, indoor, within a vehicle, in the night time, or in the rainy day, etc.), perform image acquisition and processing so as to provide the needed image information to the user.
Among many camera applications, video surveillance is one area that seizes attention in recent years, and its application scope as well as importance grows with time too. For example, the surveillance camera installed on the roadside can be provided to the traffic authority to seize the traffic condition at any time and to perform the traffic sign change or to increase the traffic guidance flexibility. The surveillance camera installed at home can let the residents seize the real time situation of their homes, especially when there is a long way travel, or there is elderly people, weak people, woman or children, that is, the residents can go for a long travel with peace of mind. In addition, the installation of the surveillance camera in the alley or the peripheral of a building can let the police or the guard seize the public security condition, especially when there is crime involved, the police can check out the video to watch its content so as to seize the condition at the crime scene and the features of the suspect.
Since crime usually occurs in the night, and the crime scene is usually at the dead space or dark corner in the community which no body pays attention to, hence, the near infrared camera that has night photography function has become necessity for video surveillance. The wavelength of the near infrared is about in the range 700 to 4,000 nm. Through the sensing of image in the near infrared wavelength range, the near infrared camera can generate clearer image in the night or in the dark place.
However, the image captured with a near infrared camera usually has the issues of insufficient contrast, insufficiently sharp features of foreground object, hence, the expected subject image is usually not clear, and its features are thus difficult to be recognized.